Red Hat AI Inference Server Model Optimization Tools 3.3.5 (CUDA) is affected by multiple security vulnerabilities identified under CVE-2025-14926 and 16 additional CVEs. These vulnerabilities involve various CWEs such as CWE-94 (Code Injection), CWE-502 (Deserialization of Untrusted Data), CWE-190 (Integer Overflow), CWE-78 (OS Command Injection), and others, indicating a broad spectrum of potential security issues. The Red Hat advisory RHSA-2026:30078 acknowledges these vulnerabilities but does not currently provide any fixes or patches. The product is not a cloud service, so remediation depends on vendor updates. No known exploits in the wild have been reported at this time.

The vulnerabilities collectively pose a high risk to systems running Red Hat AI Inference Server Model Optimization Tools 3.3.5 (CUDA). Potential impacts include unauthorized code execution, privilege escalation, denial of service, and other security breaches depending on the specific CWE involved. Since no patches are currently available, affected systems remain vulnerable until Red Hat issues a fix.

No official fix or patch is currently available for these vulnerabilities. Users should monitor Red Hat's security advisories for updates and apply patches promptly once released. Until then, consider limiting exposure of the affected software to untrusted inputs and networks to reduce risk. Follow Red Hat's guidance and contact Red Hat Product Security for further assistance if needed.